METABOLISM AND PHARMACOKINETICS OF VITAMIN D IN PATIENTS WITH CYSTIC FIBROSIS

Abstract

Patients with cystic fibrosis (CF) commonly have lower circulating concentrations of 25-hydroxyvitamin D (25(OH)D) than healthy populations. We comprehensively compared measures of vitamin D metabolism among individuals with CF and healthy control subjects. In a cross-sectional study, serum from participants with CF (N=83) and frequency-matched healthy control subjects by age and race (N=82) were analyzed for: 25(OH)D₂ and 25(OH)D₃, 1α,25-dihydroxyvitamins D₂ and D₃ (1α,25(OH)₂D₂ and 1α,25(OH)₂D₃), 24,25-dihydroxyvitamin D₃ (24,25(OH)₂D₃), 4β,25-dihydroxyvitamin D₃ (4β,25(OH)₂D₃), 25-hydroxyvitamin D₃-3-sulfate (25(OH)D₃-S), and 25-hydroxyvitamin D₃-3-glucuronide (25(OH)D₃-G). In a 56-day prospective pharmacokinetic study, ~25 μg deuterium-labeled 25(OH)D₃ (d₆-25(OH)D₃) was administered intravenously to participants (N=5 with CF, N=5 control subjects). Serum was analyzed for d₆-25(OH)D₃ and d₆-24,25(OH)₂D₃, and pharmacokinetic parameters were estimated. In the cross-sectional study, participants with CF had similar mean (SD) total 25(OH)D concentrations as control subjects (26.7 [12.3] vs. 27.7 [9.9] ng/mL) and had higher vitamin D supplement use (53% vs. 22%). However, participants with CF had lower total 1α,25(OH)₂D (43.6 [12.7] vs. 50.7 [13.0] pg/mL), 4β,25(OH)₂D₃ (52.1 [38.9] vs. 79.9 [60.2] pg/mL), and 25(OH)D₃-S (17.7 [11.6] vs. 30.1 [12.3] ng/mL) (p < 0.001 for all). The pharmacokinetics of d₆-25(OH)D₃ and d₆-24,25(OH)D₃ did not differ between groups. In summary, although 25(OH)D concentrations were comparable, participants with CF had lower 1α,25(OH)₂D, 4β,25(OH)₂D₃, and 25(OH)D₃-S concentrations than healthy controls. Neither 25(OH)D₃ clearance, nor formation of 24,25(OH)₂D₃, appears to account for these differences and alternative mechanisms for low 25(OH)D in CF (i.e., decreased formation, altered enterohepatic recirculation) should be explored.

1. Introduction

Vitamin D deficiency, defined by low circulating concentrations of 25-hydroxyvitamin D (25(OH)D), is ubiquitous in patients with cystic fibrosis (CF) (1). Most patients with this disorder require vitamin D supplementation, often at high doses, to maintain physiologically normal 25(OH)D concentrations. Vitamin D deficiency in patients with CF is associated with low bone mineral density, impaired innate and acquired immunity, and increased risk of pulmonary complications (1–3). These problems are further compounded due to decreased ability to exercise, malnutrition due to poor absorption of nutrients, and use of corticosteroids (2, 3). The Cystic Fibrosis Foundation recommends a target 25(OH)D of ≥ 30 ng/mL (75 nM) (4). However, up to 90% of persons with CF do not achieve this recommended threshold despite routine supplementation (2, 4–7), and the specific mechanisms by which their 25(OH)D concentrations remain low are poorly understood.

Vitamin D exists in two primary forms: vitamin D₂ and vitamin D₃, both of which follow similar pathways of metabolism (6, 8) (Figure S1). Derived mostly from sunlight, vitamin D is converted in the liver to 25(OH)D, the most widely used biomarker of long-term vitamin D exposure due to its long plasma half-life of 2 to 3 weeks (8, 9). 25(OH)D₃, which is more abundant than 25(OH)D₂, undergoes metabolism by cytochrome P450 (CYP) 27B1, also called 1α-hydroxylase, to form the biologically active 1α,25-dihydroxyvitamin D₃ (1α,25(OH)₂D₃) (7, 10) (Figure S1). 25(OH)D₃ may also be cleared from the blood through hydroxylation or conjugation into several inactive metabolites (11): 24R,25-dihydroxyvitamin D₃ (24,25(OH)₂D₃, the most abundant metabolite of 25(OH)D clearance) (7, 10), 1β,25-dihydroxyvitamin D₃ (12), 4β,25-dihydroxyvitamin D₃ (4β,25(OH)₂D₃) (13), 25-hydroxyvitamin D₃-3-sulfate (25(OH)D₃-S) (14), and 25-hydroxyvitamin D₃-3-glucuronide (25(OH)D₃-G) (15) (Figure S1).

Numerous mechanisms for low 25(OH)D in persons with CF have been hypothesized: reduced time outdoors, poor intestinal absorption of fat-soluble vitamins such as vitamin D, reduced capacity to hydroxylate vitamin D in the liver, reduced storage capacity due to low body fat (1), impaired enterohepatic recirculation of vitamin D metabolites enhancing their loss (18), urinary loss of vitamin D binding protein (VDBP) and albumin (the major carriers of vitamin D metabolites) (16), and increased 1α-hydroxylase expression in pulmonary epithelial cells that drive 1α,25(OH)₂D formation from 25(OH)D (17). Understanding the specific pathways that contribute to low circulating 25(OH)D concentrations in CF may improve how we assess and treat vitamin D-related complications of CF. Thus, we investigated the hypothesis that individuals with CF have altered concentrations of multiple vitamin D metabolites and increased clearance of 25(OH)D.

2. Methods

2.1. Study Design

We conducted two complementary studies to increase the understanding of vitamin D metabolism in CF. First, we compared a comprehensive set of vitamin D metabolites between participants with CF and healthy control subjects in a cross-sectional study. Second, we determined the pharmacokinetics of vitamin D metabolism by administering intravenous deuterium-labeled 25-hydroxyvitamin D₃ (d₆-25(OH)D₃) to participants with CF and healthy controls. Both studies were approved by the University of Washington Institutional Review Board. The pharmacokinetic study was approved by the US Food and Drug Administration’s Investigational New Drug program and registered on clinicaltrials.gov (NCT03104855).

2.1. Study Populations

For the cross-sectional study, we recruited participants ≥ 18 years of age with a clinical diagnosis of CF from specialty clinics at the University of Washington. All genotypes of CF were included. Participants were required to be clinically stable, defined as no hospitalization or intravenous antibiotics administered within two weeks of study visit. Healthy control subjects were identified from the Healthy Kidney Study, a cross-sectional study that previously collected biosamples and data using a similar protocol to serve as control subjects for a broad range of research studies (18, 19). Control subjects were frequency matched to CF participants by age and race.

We then invited participants with CF from the cross-sectional study to enroll in the prospective pharmacokinetic study. Healthy control subjects were recruited from the Seattle, WA area, frequency matched by age and race. For all participants in the pharmacokinetic study, inclusion criteria were age 18 years or older and a serum 25(OH)D concentration between 10–50 ng/mL. Participants were excluded if they had taken supplements of more than 400 IU/day of vitamin D₃, any vitamin D₂ supplements, 1α,25(OH)₂D₃ or analogues, calcimimetics within 3 months of enrollment, used vitamin D receptor agonists or cinacalcet within 4 weeks of enrollment, or used cytochrome P450 inhibitors or inducers within 4 weeks of enrollment. Participants on excluded drugs were allowed to washout before enrolling. Participants were also excluded for primary hyperparathyroidism, gastric bypass, tuberculosis, sarcoidosis, current pregnancy, hemoglobin < 10 g/dL, kidney transplant, or inability to provide informed consent.

2.2. Pharmacokinetic Study Design

We administered a single intravenous dose of d₆-25(OH)D₃ (26,26,26,27,27,27-d₆; ISOTEC, Stable Isotope Division, Sigma-Aldrich, Miamisburg, OH) to participants with CF and healthy controls (9). The dose of d₆-25(OH)D₃ was individualized for each study participant and calculated to target a maximum initial concentration of d₆-25(OH)D₃ of 5 ng/mL. The dose was calculated as (5 ng/mL) × (estimated blood volume), where the estimated blood volume was calculated as described by Nadler et al. (20). Blood samples were collected at 15 minutes, 4 hours, 1, 4, 7, 14, 21, 28, 42, and 56 days after d₆-25(OH)D₃ administration. Serum was harvested by centrifugation and samples were stored at −80°C until analysis.

2.3. Vitamin D Metabolite Sample Analysis

We used dedicated liquid chromatography tandem mass spectrometry assays with internal standards and external calibration to quantify 25(OH)D₂, 25(OH)D₃, 1α,25(OH)₂D₂, 1α,25(OH)₂D₃, 24,25(OH)₂D₃, 4β,25(OH)₂D₃, 25(OH)D₃-S, and 25(OH)D₃-G (supplemental material). Total 25(OH)D and 1α,25(OH)₂D were calculated by summing the concentrations of 25(OH)D₂ and 25(OH)D₃, or 1α,25(OH)₂D₂ and 1α,25(OH)₂D₃, respectively.

Free (i.e., unbound) concentrations of 25(OH)D₃ and 1α,25(OH)₂D₃ were calculated using the previously published equation (21, 22),

$${\text{D}}_{\text{free}}=\frac{{\text{D}}_{\text{total}}}{(1+({\text{k}}_{\text{ALB}}\cdot [{\text{ALB]) + (k}}_{\text{VDPB}}\cdot [\text{VDBP]))}}$$

where D_free is the calculated free concentration of 25(OH)D₃ or 1α,25(OH)₂D₃. D_total is the measured serum concentration of 25(OH)D₃ or 1α,25(OH)₂D₃. [ALB] and [VDBP] are the measured serum albumin and VDBP concentrations, respectively. k_ALB and k_VDPB are the association constants for albumin and VDBP, respectively (21–23). The values of the 25(OH)D₃ association constants for albumin and VDBP are 6 × 10⁵ M⁻¹ and 7 × 10⁸ M⁻¹, respectively. The values of the 1α,25(OH)₂D₃ association constants for albumin and VDBP are 5.4 × 10⁴ M⁻¹ and 3.7 × 10⁷ M⁻¹, respectively. The percent unbound of 25(OH)D₃ and 1α,25(OH)₂D₃ was estimated using % unbound = D_free/D_total ·100%.

For the pharmacokinetic study, the liquid chromatography-mass spectrometry assays were adapted to quantify d₆-25(OH)D₃ and deuterium-labeled 24,25-dihydroxyvitamin D₃ (d₆-24,25(OH)₂D₃) in the serum samples (supplemental materials). In preliminary experiments, no other deuterated metabolites following d₆-25(OH)D₃ administration were detected (data not shown).

2.4. Pharmacokinetic Data Analysis

Serum concentrations of d₆-25(OH)D₃ and d₆-24,25(OH)₂D₃ were analyzed using non-compartmental analysis with Phoenix WinNonlin software (version 8.0.0). Due to variability in the dose administered, the concentration-time profiles were dose-normalized and scaled to a 25 μg dose of d₆-25(OH)D₃. For d₆-25(OH)D₃, C_max and T_max were the observed dose-normalized peak d₆-25(OH)D₃ concentration and the time when the peak concentration was observed, respectively. The elimination rate constant (k) was estimated using uniform weighting of timepoints after 14 days. Elimination half-life of d₆-25(OH)D₃ was calculated as t_1/2 = ln(2)/k. Area under the concentration-time curve (AUC_0−∞) was calculated using the trapezoidal rule to calculate AUC_0-last (to the last observed timepoint) and extrapolated to time = infinity. The percent of AUC extrapolated was calculated as AUC_0-last/AUC_0−∞ × 100. The systemic clearance (CL) of d₆-25(OH)D₃ was calculated as dose/AUC_0−∞. The steady-state volume of distribution (V_ss) was calculated as (dose∙AUMC)/(AUC)², where the AUMC was calculated as the area under the first moments curve.

Concentrations of d₆-24,25(OH)₂D₃ were not dose-normalized for pharmacokinetic analyses. The elimination rate constant of d₆-24,25(OH)₂D₃ was not estimated as measured concentrations were near the lower limit of quantification resulting in too few timepoints describing decreasing d₆-24,25(OH)₂D₃ concentrations. The molar AUC_0-last ratio was calculated as (d₆-24,25(OH)₂D₃ AUC_0-last)/(d₆-25(OH)D₃ AUC_0-last) using the observed d₆-25(OH)D₃ AUC_0-last (not dose-normalized).

2.5. Statistical Analysis

We estimated the difference in concentrations of each vitamin D metabolite and other biomarkers between participants with CF and healthy controls using linear regression. Adjusted models included the variables of age, sex, and season or age, sex, season, and upstream vitamin D metabolite: total 25(OH)D for 1α,25(OH)₂D, 25(OH)D₃ for 25(OH)D₃-derived metabolites, or 25(OH)D₂ for 1α,25(OH)D₂.

Pharmacokinetic parameters of participants with CF were compared to healthy controls using a Mann-Whitney U test. A value of p < 0.05 was considered statistically significant.

3. Results

3.1. Participant Characteristics

In total, 83 participants with CF and 82 healthy controls were enrolled (Table 1). The mean age for both groups was similar, but the percentage of males in the healthy controls was lower than that of the CF group (38% vs. 55%). More than 85% of participants with CF had a clinical diagnosis of pancreatic insufficiency, and 53% were receiving vitamin D supplementation compared with 22% of healthy controls. Approximately 50% and 40% of the participants with CF were homozygous and heterozygous for the cystic fibrosis transmembrane conductance regulator (CFTR) ΔF508 gene variation, respectively.

Table 1.

Characteristics of study participants

	Cross-Sectional Study		Pharmacokinetic Study
	Participants with CF (n = 83)	Healthy Controls (n = 82)	Participants with CF (n = 5)	Healthy Controls (n = 5)
Age (years)		32.5 ± 11.2		24.8 ± 3.3
Male	45 (55%)	31 (38%)	4 (80%)	1 (20%)
BMI (kg/m2)	23.0 ± 3.4	25.5 ± 4.9	25.0 ± 3.1	22.4 ± 3.0
Race
White	80 (96%)	79 (96%)	5 (100%)	5 (100%)
Other	3 (4%)	3 (4%)	0 (0%)	0 (0%)
Season of Sample Collection
Winter	23 (37%)	37 (45%)	2 (40%)	0 (0%)
Spring	14 (17%)	31 (38%)	1 (20%)	4 (80%)
Summer	11 (13%)	10 (12%)	2 (40%)	1 (20%)
Fall	35 (42%)	4 (5%)	0 (0%)	0 (0%)
Pancreatic Insufficient	71 (86%)	--a	3 (60%)	--
Vitamin D Supplementation	44 (53%)	18 (22%)	--	--
FGF (pg/mL)	44.4 ± 18.3	41.4 ± 11.8	--	--
PTH (pg/mL)	43.1 ± 19.0	40.6 ± 15.4	--	--
CFTR Genotype
Homozygous ΔF508	33 (40%)	--	1 (20%)	--
Heterozygous ΔF508	42 (51%)	--	3 (60%)	--
Other	8 (10%)	--	1 (20%)	--

Results are reported as mean ± SD or count (%)

^a-- Not determined

3.2. Vitamin D Metabolite Concentrations in the Cross-Sectional Study

Only 35% of participants with CF had total 25(OH)D concentrations ≥ 30 ng/mL. Sixty-seven percent and 78% of participants with CF and healthy controls had total 25(OH)D concentrations ≥ 20 ng/mL, respectively. Mean ± standard deviation concentrations of 25(OH)D were comparable between participants with CF and healthy controls (26.7 ± 12.3 ng/mL and 27.7 ± 9.9 ng/mL, respectively; Table 2). Total 1α,25(OH)₂D was systematically lower in participants with CF than in healthy controls, which persisted after adjustment for age, sex, season and total 25(OH)D (mean difference: −7.1 pg/mL; 95% CI: −11.0, −3.2; p < 0.001). Similar results were observed for 4β,25(OH)₂D₃ and 25(OH)D₃-S. Concentrations of 24,25(OH)₂D₃ and of 25(OH)D₃-G did not differ between the two groups before or after adjusted analyses.

Table 2:

Cross-sectional concentrations of 25-hydroxyvitamin D, its metabolites, and vitamin D binding proteins in participants with cystic fibrosis and healthy controls

	Unadjusted Data				Adjusted for age, sex, and season		Adjusted for age, sex, season, and 25(OH)D-specific metabolite
	Participants with CF (n = 83)	Healthy Controls (n = 82)	CF vs. HealthyDifference (95% CI)	p-value	CF vs. Healthy Adjusted Difference (95% CI)	p-value	CF vs. Healthy Adjusted Difference (95% CI)	p-value
Total Vitamin D Metabolites
25(OH)D (ng/mL)	26.7 ± 12.3	27.7 ± 9.9	−1.0 (−4.4, 2.4)	0.57	1.2 (−2.2, 4.6)	0.48	--a	--
1α,25(OH)2D (pg/mL)	43.6 ± 12.7	50.7 ± 13.0	−7.1 (−11.0, −3.2)	<0.001	−7.8 (−11.9, −3.6)	<0.001	−7.9 (−12.0, −3.7)	<0.001
Vitamin D3 Metabolites
25(OH)D3 (ng/mL)	25.3 ± 12.0	26.8 ± 10.1	−1.5 (−4.8, 1.9)	0.39	0.6 (−2.8, 4.0)	0.73	--	--
Unbound 25(OH)D3 (pg/mL)b	17.9 ± 8.4	18.2 ± 6.8	−0.3 (−2.7, 2.1)	0.80	--	--	--	--
% Unbound 25(OH)D3	0.072 ± 0.009	0.06 9 ± 0.00 1	0.0023 (0.0003, 0.0043)	0.027	--	--	--	--
1α,25(OH)2D3 (pg/mL)	42.3 ± 13.3	49.5 ± 12.8	−7.2 (−11.1, −3.2)	<0.001	−7.9 (−12.2, −3.6)	<0.001	−8.0 (−12.2, −3.7)	<0.001
Unbound 1α,25(OH)2D3 (pg/mL)b	0.51 ± 0.15	0.57 ± 0.13	−0.058 (0.100, 0.015)	0.009	--	--	--	--
% Unbound 1α,25(OH)2D3	0.50 ± 0.06	0.49 ± 0.07	0.018 (0.001, 0.037)	0.070	--	--	--	--
24,25(OH)2D3 (ng/mL)	1.7 ± 1.2	1.7 ± 1.0	0.0 (−0.4, 0.3)	0.81	0.1 (−0.2, 0.5)	0.51	0.1 (−0.1, 0.2)	0.44
4β,25(OH)2D3 (pg/mL)c	52.1 ± 38.9	79.8 ± 60.5	−27.6 (−43.2, −12.0)	<0.001	−25.1 (−40.8, −9.3)	0.002	−26.6 (−40.3, −13.0)	<0.001
25(OH)D3−S (ng/mL)	17.7 ± 11.6	30.1 ± 12.3	−12.4 (−16.0, −8.7)	<0.001	−11.2 (−14.9, −7.5)	<0.001	−11.7 (−14.4, −8.9)	<0.001
25(OH)D3−G (ng/mL)	1.9 ± 1.2	2.2 ± 1.6	−0.3 (−0.7, 0.1)	0.13	−0.2 (−0.6, 0.2)	0.37	−0.2 (−0.5, 0.0)	0.09
Vitamin D2 Metabolites
25(OH)D2 (ng/mL)	1.3 ± 5.4	0.9 ± 1.9	0.4 (−0.8, 1.7)	0.49	0.6 (−1.0, 2.2)	0.44	--	--
1α,25(OH)2D2 (pg/mL)	1.3 ± 4.2	1.2 ± 2.5	0.1 (−0.9, 1.1)	0.87	0.1 (−1.2, 1.4)	0.86	−0.4 (−0.8, 0.1)	0.10
Vitamin D Binding Proteins
VDBP (μg/mL)	264 ± 41	274 ± 56	−10 (–25, 5)	0.19	--	--	--	--
Albumin (g/dL)	3.9 ± 0.4	4.3 ± 0.3	−0.4 (−0.5, −0.3)	<0.001	--	--	--	--

Results are reported as mean ± SD for unadjusted data and difference (95% CI) for adjusted models

^a-- Not determined

^bUnbound concentrations of 25(OH)D₃ and 1α,25(OH)₂D₃ were calculated using previously published equations (21, 22)

^cFor the CF group, n = 82 due to insufficient sample volume for one study participant

Serum VDBP concentrations were similar between the two groups, while serum albumin concentrations were lower in participants with CF than in healthy controls (Table 2). Although unbound 25(OH)D₃ concentrations were comparable between participants with CF and healthy controls, participants with CF had a higher percent of unbound 25(OH)D₃ than healthy controls (mean difference: 0.0023; 95% CI: 0.0003, 0.0043). Unbound 1α,25(OH)₂D₃ concentrations were 10% lower in participants with CF than in healthy controls (0.508 ± 0.151 vs. 0.566 ± 0.126 pg/mL; p < 0.05), while the percent of unbound 1α,25(OH)₂D₃ was comparable between the two groups.

3.3. Pharmacokinetic Study Participants

A total of 5 participants with CF and 5 healthy controls participated in the pharmacokinetic study (Table 1). Participants with CF were older and had a higher percentage of males than the healthy control group.

3.4. Pharmacokinetics of d₆-25-hydroxyvitamin D₃

The d₆-25(OH)D₃ and d₆-24,25(OH)₂D₃ concentration vs. time profiles and pharmacokinetic parameters are presented in Figure 1 and Table 3, respectively. The peak concentration (C_max) of d₆-25(OH)D₃ was observed at the earliest timepoint (15 min) for all study participants, after which d₆-25(OH)D₃ declined over time. No differences were noted in the mean clearance (CL), steady-state volume of distribution (V_ss), or terminal half-life (t_1/2) of d₆-25(OH)D₃ between participants with CF and healthy controls (p > 0.05 for all). The coefficient of variation (CV%) for d₆-25(OH)D₃ pharmacokinetic parameters was less than 20% for both groups.

Figure 1.

Deuterated 25-hydroxyvitamin D₃ and 24,25-dihydroxyvitamin D₃ concentrations were similar between CF and healthy controls over time. d₆-25(OH)D₃ concentrations, dose-adjusted to 25 μg d₆-25(OH)D₃, are presented as circles and solid lines, while d₆-24,25(OH)₂D₃ concentrations are presented as squares and dashed lines. Group means are presented for participants with CF (closed symbols) and healthy controls (open symbols). The vertical whiskers indicate the standard deviation in each direction.

Table 3:

Summary of pharmacokinetic (PK) parameters after intravenous administration of d₆-25-hydroxyvitamin D₃ in participants with cystic fibrosis and healthy controls

Parametera	Participants with CF (n=5)	Healthy Controls (n=5)	p-value
Actual dose d6-25(OH)D3 (μg)	26.4 ± 3.6	22.0 ± 3.2	0.11
d6-25(OH)D3 Dose-Adjusted PK Parametersb
Cmax (ng/mL)	8.8 ± 1.4	9.4 ± 1.9	0.84
AUC0-last (ng·day/mL)	58.3 ± 9.7	67.2 ± 8.3	0.22
AUC0−∞ (ng·day/mL)	65.1 ± 12.5	73.9 ± 9.0	0.31
Extrapolated AUC (%)	10.0 ± 3.6	8.9 ± 1.6	0.69
CL (mL/day)	397 ± 73	342 ± 41	0.31
VSS (L)	8.4 ± 1.4	7.2 ± 1.1	0.10
t1/2 (day)	16.2 ± 3.3	15.8 ± 1.6	0.84
d6-24,25(OH)2D3 PK Parameters
Cmax (ng/mL)	0.16 ± 0.02	0.13 ± 0.02	0.07
AUC0-last (ng·day/mL)	5.9 ± 1.2	4.8 ± 0.9	0.15

^aResults are reported as mean ± SD

^bPharmacokinetic parameters for d₆-25(OH)D₃ were calculated using dose-normalized d₆-25(OH)D₃ serum concentrations adjusted to 25 μg d₆-25(OH)D₃.

The mean C_max of d₆-24,25(OH)₂D₃ was comparable between participants with CF and healthy controls and peaked at a median (interquartile range) of 14 (4–14) days for both groups. The d₆-24,25(OH)₂D₃ AUC_0-last and molar AUC ratio did not differ between participants with CF and healthy controls. Due to variability in the d₆-24,25(OH)₂D₃ data, the terminal half-life and AUC_0−∞ could not be estimated.

4. Discussion

The present investigation is one of the first to evaluate a comprehensive set of vitamin D metabolites and the pharmacokinetics of 25(OH)D in adults with CF. Only 35% of participants with CF had total 25(OH)D concentrations greater than or equal to the guideline-recommended threshold of 30 ng/mL despite over 50% of them taking vitamin D supplements. Of the metabolites studied, concentrations of total 1α,25(OH)₂D, 4β,25(OH)₂D₃, and 25(OH)D₃-S were lower in participants with CF than in healthy controls, even after covariate adjustment. There were no differences in the pharmacokinetics of d₆-25(OH)D₃ or in the formation of d₆-24,25(OH)₂D₃ between the two groups.

25(OH)D is the most studied vitamin D metabolite in CF, and our results are consistent with that of others reporting low serum concentrations despite routine vitamin D supplementation (2, 4–7). The downstream metabolites of 25(OH)D, however, have received significantly less attention. Greer et al. reported that total 1α,25(OH)₂D concentrations were lower in children, adolescents, and adults with CF than in healthy controls despite comparable 25(OH)D concentrations, similar to our findings in adults (24). Weisman et al. reported no differences in 24,25(OH)₂D₃ concentrations between children with and without CF (2.2 ng/mL ± 0.6 and 2.6 ± 1.3 ng/mL, respectively), but observed concentrations higher than seen in our study, possibly due to differences in analytical technique or study population (25). To our knowledge, this is the first study to assess 24,25(OH)₂D₃, 4β,25(OH)₂D₃, 25(OH)D₃-G and 25(OH)D₃-S concentrations in adults with CF.

In a mechanism thought to protect against vitamin D toxicity, 25(OH)D₃ is metabolized by CYP24A1 into 24,25(OH)₂D₃, the predominant downstream metabolite of 25(OH)D₃ in the serum (11, 18). We hypothesized that increased 25(OH)D₃ clearance and 24,25(OH)₂D₃ formation could contribute to low 25(OH)D concentrations in CF, but did not observe significant differences in these pharmacokinetic parameters between participants with CF and healthy controls. While our study was likely underpowered to detect more modest differences in pharmacokinetic parameters between the two groups, it seems reasonable to conclude that altered 25(OH)D₃ clearance does not contribute substantially to low 25(OH)D concentrations in persons with CF.

The lower concentrations of 25(OH)D₃-S in participants with CF than in healthy controls offer some insight into potential mechanisms for low 25(OH)D in CF. In the liver, 25(OH)D₃ undergoes sulfation by sulfotransferase 2A1 (SULT2A1) into 25(OH)D₃-S, which has been hypothesized to act as a vitamin D reservoir as it is protected from further hydroxylation into 24,25(OH)₂D₃ or 1,25(OH)₂D₃ unlike 25(OH)D₃ (26). 25(OH)D₃-S may be excreted into bile, reabsorbed in the small intestine, and deconjugated back into 25(OH)D₃ (14, 27). Excess biliary excretion or impaired enterohepatic recirculation of 25(OH)D₃-S may therefore result in low concentrations of 25(OH)D₃-S and 25(OH)D₃ in CF and requires further investigation.

It is less clear how our findings of low 1α,25(OH)₂D and 4β,25(OH)₂D₃ in participants with CF explain low 25(OH)D in CF, as they suggest reduced formation of these metabolites from 25(OH)D. Nonetheless, these observations may offer some insight on the pathophysiology of their respective pathways in CF. As the biologically active form of vitamin D, 1α,25(OH)₂D concentrations are tightly regulated in healthy persons, in particular by fibroblast growth factor-23 (FGF-23) and parathyroid hormone (PTH). Given that concentrations of FGF-23 and PTH were similar between participants with CF and healthy controls, other factors may be responsible for reduced 1α-hydroxylase activity in CF, including defects to the CFTR, which is abundantly expressed in the kidney (28). Similarly, patients with CF may experience down-regulation of hepatic CYP3A4, the enzyme responsible for the formation of 4β,25(OH)₂D₃ (13). Whether these changes are related to the pathology of CF, interactions with medications (ivacaftor, for example, is also metabolized by CYP3A4 (29)), or other mechanisms remain to be explored.

Strengths of this study include the comprehensive evaluation of vitamin D metabolites using precise and accurate measurements, and the first-ever use of gold-standard pharmacokinetic methods to determine 25(OH)D₃ clearance in CF. There are a few limitations. First, due to the cross-sectional nature of the study and the lack of clinical outcomes, we could not determine whether the measured differences in metabolites of 25(OH)D have clinical implications. Second, we did not directly measure enzymatic activity. It is possible that low concentrations of 1α,25(OH)₂D, 4β,25(OH)₂D₃, and 25(OH)D₃-S are not due to reduced formation, but rather increased elimination. Third, we did not measure 3-epi-25-hydroxyvitamin D₃, an increasingly recognized epimer of 25(OH)D₃, which circulates at approximately 4–5% of 25(OH)D₃ concentrations (30, 31). As we found measured 25(OH)D₃ clearance to be similar between CF and controls, potential differences in C3 epimerization of 25(OH)D₃ seem insufficient to explain low 25(OH)D in CF, though the importance of C3 epimers of vitamin D in CF remains to be determined. Fourth, we did not measure vitamin D₃, the precursor to 25(OH)D₃, and cannot make any inferences with regards to this metabolic pathway.

5. Conclusions

Compared with healthy participants, those with CF had lower total 1α,25(OH)₂D, 4β,25(OH)₂D₃ and 25(OH)D₃-S, perhaps as a consequence of decreased formation in CF. As there were no differences in the systemic clearance or half-life of d₆-25(OH)D₃ between our two study groups, increased 25(OH)D₃ elimination or formation of 24,25(OH)₂D₃ are unlikely to contribute to low 25(OH)D concentrations in persons with CF. Other factors that may explain 25(OH)D deficiency in CF, such as altered absorption of vitamin D, decreased hepatic 25-hydroxylation, or losses through decreased enterohepatic recirculation need further investigation. Identifying the precise mechanisms involved may improve how we assess and treat vitamin D-related complications of CF.

HIGHLIGHTS.

• Assessed 8 vitamin D metabolites in participants with cystic fibrosis and controls

• Total 25(OH)D was comparable between participants with cystic fibrosis and controls

• Total 1α,25(OH)₂D, 4β,25(OH)₂D₃, 25(OH)D₃-3-sulfate were lower in cystic fibrosis

• Clearance of 25(OH)D₃ did not differ between groups in a pharmacokinetic study

Abbreviations:

1α,25(OH)₂D

1α,25-dihydroxyvitamin D

1α,25(OH)₂D₂

1α,25-dihydroxyvitamin₂

1α,25(OH)₂D₃

1α,25-dihydroxyvitamin D₃

24,25(OH)₂D₃

24,25-dihydroxyvitamin D₃

25(OH)D

25-hydroxyvitamin D

25(OH)D₂

25-hydroxyvitamin D₂

25(OH)D₃

25-hydroxyvitamin D₃

25(OH)D₃-G

25-hydroxyvitamin D₃-3-glucuronide

25(OH)D₃-S

25-hydroxyvitamin D₃-3sulfate

4β,25(OH)₂D₃

4β,25-dihydroxyvitamin D₃

AUC_0−∞

area under the concentration-time curve

AUC_0-last

area under the concentration-time curve to the last observed time point

CFTR

cystic fibrosis transmembrane conductance regulator gene

CL

systemic clearance

C_max

maximum observed concentration

CF

cystic fibrosis

d₆-24,25(OH)₂D₃

deuterium-labeled 24,25-dihydroxyvitamin D₃

d₆-25(OH)D₃

deuterium-labeled 25-hydroxyvitamin D₃

FGF-23

fibroblast growth factor-23

PK

pharmacokinetic

PTH

parathyroid hormone

SULT2A1

sulfotransferase 2A1

t_1/2

elimination half-life

T_max

time of maximum concentration

VDBP

vitamin D binding protein

V_ss

steady-state volume of distribution